Polymarket US Historical Data

Polymarket US order books and trades as daily Parquet files, collected from the Polymarket US WebSocket with nanosecond timestamps. Data from 2026-09-10.

2
data types
daily
Parquet file per market
600
free downloads
End point
/v1/polymarket_us/l2/{YYYY/MM/DD}/{market_slug}.parquet
/v1/polymarket_us/trades/{YYYY/MM/DD}/{market_slug}.parquet

Download with Python

import pandas as pd

market_slug = "aec-nfl-pit-cle-2026-10-01"  # Polymarket US market slug
date = "2026/10/01"  # UTC date, YYYY/MM/DD

url = f"https://api.pmdata.dev/v1/polymarket_us/l2/{date}/{market_slug}.parquet"
# trades_url = f"https://api.pmdata.dev/v1/polymarket_us/trades/{date}/{market_slug}.parquet"

df = pd.read_parquet(url, storage_options={"api_key": "<YOUR_API_KEY>"})
# timestamp and local_timestamp load as datetime64[ns]; nanoseconds are kept
print(df.head())

Pricing

Free: 600 downloads at sign-up, then 50 per month, for data older than 3 weeks.

Plus: Polymarket US: $49/month for 1M Polymarket US downloads and data from the last 3 weeks. It is separate from Plus: Polymarket and Plus: Predict.fun; hold any of them.

See pricing

Frequently asked questions